11 thoughts on “Lockheed Martin’s Fusion Plans”

  1. I put fusion projects into three categories. Large scale hot fusion, small scale hot fusion, and condensed matter low energy fusion.

    When I was in college, the claim was 20 years to commercial power plants using large scale hot fusion. 40 years later they are no closer, but essentially all of DOE’s research budget goes to large scale hot fusion. Big government likes big research projects.

    Then, there’s the various descendents of the Farnsworth Fusor and Polywell. which I personally collectively call table top hot fusion. There’s not much government research funding, but there are some credible privately funded efforts, such as this Lockheed announcement.

    Cold fusion is still shuffling forward despite the criticism from mainstream physicists. There are literally hundreds of repeatable experiments done in dozens of labs showing there’s *something* there. The annual ICCF conference includes lots of papers on test results from around the world. It’s a tough read, but I recommend Ed Storms collection of papers, results, and attempted explanations in his book “Explanation of Low Energy Nuclear reaction” The reason it’s called ‘low energy’ is not so much because the units are small, but because the energy release per nuclear reaction is 1 or 2 orders of magnitude less than DD or DT reactions in hot fusion.

    At least one of the above 3 concepts is going to commercially succeed in making electricity production, and I enjoy watching all of them.

    1. I’m still not sure if anything is going to succeed commercially, but the odds are raised quite a bit the more things get tried. All of what you’re calling “hot fusion” ultimately comes down to trying out various fairly expensive combinations of some magnetic field configuration, maybe a giant laser or two, and a lot of high voltage. (I guess you have to add “steampunk” for the General Fusion guys…). The ITER/tokamak combination has a very high probability of getting to ignition, but its probability of commercial viability is no higher than a lot of the fly-by-nights. So it’s a lot more likely that you’ll get a winner if you’re rolling the dice 20 times than just once. The policy and investment guys finally seem to be waking up to this fact.

      The problem with all of the approaches is that each one requires a lot of capital and, once the capital is raised, the construction of exotic, bug-prone experimental apparatus. That makes them take a long time. If we had the ability to do decent simulations of plasma systems, things would go a lot faster. I keep wondering whether there’s some threshold in particle-in-cell simulation granularity where the sims start agreeing with experiment so well that you can suddenly do thousands of different configurations to find two or three that have a very high probability of being viable.

  2. If there is any there, there, this will be huge. It should have obvious space potential even if not in zero g (but here’s hoping for that as well.)

    Why would they announce so early?

    1. “Why would they announce so early?”

      That is an excellent question, and one to which I haven’t heard a cogent answer. Not a very skunkworks-esque thing to do, is it? The conventional wisdom seems to be that they’re trying to attract funding, but that’s kind of a bizarro explanation for something happening in a big aerospace conglomerate. I suppose that they could be trying to get ARPA-E to shovel some cash at them, but private funding would be weird. Here are my guesses:

      1) The aforementioned ARPA-E gambit.

      2) McGuire’s group went open-loop and mounted a PR blitz to forestall some attempt by LM management to scrap the program.

      3) LM might be thinking about spinning the thing off and they’re looking for private equity sugardaddies.

      4) They’ve really got something and they’re doing PR prep for a much more earthshaking set of announcements.

      5) They were freaking out on having no patent apps filed and figured it would be weird for the stockholders if they filed without accompanying PR.

      6) They were worried about losing share-of-mind in the community, given high-profile announcements by the MagLIF group and the U Washington folks.

      Beyond that, I’ve got nothing. This thing’s very compact, but it’s got the same neutron embrittlement problems as everybody else (with the notable exception of General Fusion). There’s been some speculation that beta might be so high that aneutronic fuels would be possible, but I’m having a failure of imagination on that one. Still, it’s all kinda cool.

        1. McGuire is making a bit of an extraordinary claim when he says that the configuration is “inherently stable”. Presumably there are weasel words that go with that statement, but imagine what happens if you can not only have high beta (plasma pressure/magnetic pressure) but also high absolute pressure.

          That might be the secret sauce for D-He3. Of course, you still need to get the He3 from somewhere, but if you can do D-He3, you can do D-D, which breeds plenty of He3, albeit with much higher neutronicity. I could see a terrestrial D-D power industry that provided fuel for D-He3 space applications. I remain intensely skeptical of the economics of tritium breeding for terrestrial power applications, so D-D would be handy in that regard, too.

          It’s fun to salivate at the vaporware, isn’t it?

  3. I think it is neat Lockheed is putting some money into this. Even if it does not work they will learn things about dealing with high temperature plasmas and that was other applications as well.

  4. Dunno, it is so like “why are we talking about Space Elevators when the materials tech to build such a thing could more easily build a fleet of durable SSTO rockets?”

    Why are we talking fusion when a fraction of the effort could get you molton salt thorium fission reactors, subcritical particle accelerator driven fission, etc.?

    OK, so fission means nuclear proliferation and this thingy offering gobs of high energy neutrons long before it is cost effective for power generation is not? This thing means no China Syndrome when its cost is orders of magnitude more than the passive-cooled fission plants? It means no “heavy” nuclear waste when this thing is plastering the first wall with neutrons to make the durability of waste-transmuting fast fission reactors a piece of cake?

    Why are we doing this instead of a full-court press of next-gen fission?

    1. IMO, fusion is much more of a proliferation problem than a thorium breeder. U-238 + fast neutrons = lots and lots of easily separable Pu-239. Thorium breeders make lots of U-233, but it’s very hard to handle and separate.

      However, fusion has one huge advantage over thorium breeders: The chance of an accident that contaminates hundreds of square kilometers is provably zero. That’s never going to be true for a fission nuke. Sure, LFTRs are inherently safer than light water reactors, but you can’t get the risk to zero. You’re always going to be dealing with the low probability/high cost accident scenario. BTW, I once worked out that with Chernobyl, Kyshtym, and Fukushima, the existing terrestrial nukes take an average of 500 sq km of land out of service per year. That’s a tiny amount compared to most other energy sources, but it’s still a scary number. Fusion simply doesn’t have that risk.

      Second, while I am a huge fan of LFTR-style thorium breeders, fission technology takes a long, long time to develop. Yes, I know about the MSRE at Oak Ridge, but that thing:

      a) Was a toy–something like 9 MWt.

      b) Didn’t breed thorium–they added U-233 as needed.

      c) Didn’t have the kind of closed loop extraction and reprocessing that you’d need for a viable power plant.

      d) Had only scratched the surface on the metallurgy and chemistry issues needed to assure reliability.

      My guess on the best you could do with a crash LFTR program would be 20 years. That sounds like a pretty good thing to do, but 20 years is a long time.

      On the other hand, because of the safety profile for fusion, you could see the technology go from proof-of-concept to deployable technology much faster than that. Unlike fission nukes, a first generation fusion nuke doesn’t have to have every single last bug worked out to be a decent power source. You can have design iteration cycles that are measured in 3-5 years, instead of 20-30 years for fission nukes.

Comments are closed.